Arthur Zakirov wrote:
> I've planned only to improve the documentation a little. Also it seems I
> should change 0004 part, I found that extension upgrade scripts may be made
> in wrong way.
I've attached new version of the patch. In this version I removed
0004-Update-tmplinit-arguments-v6.patch. In my opinion it handled
extensions upgrade in wrong way. If I'm not mistaken currently there is
no way to upgrade a template's init function signature. And I didn't
find way to change init_method(internal) to init_method(internal,
internal) within an extension's upgrade script.
Therefore I added 0002-Change-tmplinit-argument-v7.patch. Now
DictInitData struct is passed in a template's init method. It contains
necessary data: dictoptions and dictid. And there is no need to change
the method's signature.
Other parts of the patch are same, except that they use DictInitData
structure now.
On Mon, Mar 19, 2018 at 01:52:41AM +0100, Tomas Vondra wrote:
> I wonder how much of this patch would be affected by the switch from dsm
> to mmap? I guess the memory limit would get mostly irrelevant (mmap
> would rely on the OS to page the memory in/out depending on memory
> pressure), and so would the UNLOAD/RELOAD commands (because each backend
> would do it's own mmap).
I beleive mmap requires completely rewrite 0003 part of the patch and a
little changes in 0005.
> In any case, I suggest to polish the dsm-based patch, and see if we can
> get that one into PG11.
Yes we have more time in future commitfests if dsm-based patch won't be
approved.
--
Arthur Zakirov
Postgres Professional: http://www.postgrespro.com
Russian Postgres Company